Developmental anomalies of the brain, including mental retardation and malformations, occur in approximately 2% of all liveborn children and epilepsy in about 0.5%, placing them among the most common childhood disorders known. Despite the high frequency, the molecular and cellular basis for some of the underlying disorders has been elucidated only recently, while the basis of many more remains unknown. Mutations in the transcription factor ARX have recently been described in several children with early childhood epilepsy and mental retardation, both with and without associated brain malformations. We anticipate that mutations of this gene will prove to be a relatively common cause of mental retardation and infantile epilepsy based on the wide spectrum of severity already apparent in this group of children and a recurrent mechanism for mutation in at least two of the four polyalanine tracts found in the gene. The developmental mechanism by which ARX mutations result in this wide spectrum of developmental problems is incompletely understood, although emerging data implicate disturbances in radial and nonradial cell migration, two pathways required for normal brain development. Based on preliminary data from humans and mice, the following hypotheses have been generated: (1) the type of ARX mutation predicts the phenotype in both hemizygous males and heterozygous females; (2) Arx is necessary for normal radial cell migration of projection neurons from the neocortical ventricular zone, and for nonradial migration of inhibitory interneurons and development of the basal ganglia from the ganglionic eminence; and (3) 5' polyalanine expansions of Arx seen in some patients with infantile seizures and mental retardation cause Arx aggregation in the nucleus, which results in loss of Arx function or renders the protein toxic to vulnerable neurons. To test these hypotheses, a series of experiments are proposed that will discover the mutation types in a large series of male and female patients with candidate phenotypes. In addition, a mouse will be generated with LoxP sites surrounding exon 2, allowing for the specific deletion of Arx in either the neocortex or the ganglionic eminence. Finally, mice with an expansion of a polyalanine track, mimicking a common human mutation of ARX will be constructed. The phenotypes in humans and mutant mice with different. ARX/Arx mutations will be analyzed and compared to each other and to overlapping phenotypes caused by mutations of related genes. These studies are expected to provide a greater understanding of how Arx functions in normal and abnormal development, and will contribute to our understanding of the pathogenesis of such common disorders in children as mental retardation, epilepsy, and structural anomalies of the brain.